Interaction of Functionalized Naphthalenophanes with Abasic Sites in DNA: DNA Cleavage, DNA Cleavage Inhibition, and Formation of Ligand–DNA Adducts

Article abstract of DOI:10.1002/chem.201805555

Ligands interacting with abasic (AP) sites in DNA may generate roadblocks in base-excision DNA repair (BER) due to indirect inhibition of DNA repair enzymes (e.g., APE1) and/or formation of toxic byproducts, resulting from ligand-induced strand cleavage or covalent cross-links. Herein, a series of 12 putative AP-site ligands, sharing the common naphthalenophane scaffold, but endowed with a variety of substituents, have been prepared and systematically studied. The results demonstrate that most naphthalenophanes bind to AP sites in DNA and inhibit the APE1-induced hydrolysis of the latter in vitro. Remarkably, their APE1 inhibitory activity, as characterized by IC₅₀ and K_I values, can be directly related to their affinity and selectivity to AP sites, as assessed by means of fluorescence melting experiments. On the other hand, the molecular design of naphthalenophanes has a crucial influence on their intrinsic AP-site cleavage activity (i.e., ligand-catalyzed β- and β,δ-elimination reactions at the AP site), as illustrated by the compounds either having an exceptionally high AP-site cleavage activity (e.g., 2,7-BisNP-S, 125-fold more efficacious than spermine) or being totally devoid of this activity (four compounds). Finally, the unprecedented formation of a stable covalent DNA adduct upon reaction of one ligand (2,7-BisNP-NH) with its own product of the AP-site cleavage is revealed.

Full text of DOI:10.1002/chem.201805555

A Journal of
Accepted Article
Title: Interaction of Functionalized Naphthalenophanes with Abasic
Sites in DNA: DNA Cleavage, DNA Cleavage Inhibition, and
Formation of Ligand–DNA Adducts
Authors: Coralie Caron, Xuan N. T. Duong, Régis Guillot, Sophie
Bombard, and Anton Granzhan
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201805555
Link to VoR: http://dx.doi.org/10.1002/chem.201805555
Supported by

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
Interaction of Functionalized Naphthalenophanes with Abasic
Sites in DNA: DNA Cleavage, DNA Cleavage Inhibition, and
Formation of Ligand–DNA Adducts
Coralie Caron,[a,b] Xuan N. T. Duong,[a,b] Régis Guillot,[c] Sophie Bombard,[a,b] and Anton Granzhan*[a,b]
Abstract: Ligands interacting with abasic (AP) sites in DNA may
generate roadblocks in base-excision DNA repair (BER) due to
indirect inhibition of DNA repair enzymes (e.g., APE1) and/or
formation of toxic byproducts, resulting from ligand-induced strand
cleavage or covalent cross-links. Herein, we prepared and syste-
matically studied a series of 12 putative AP-site ligands, sharing the
common naphthalenophane scaffold but endowed with a variety of
substituents. Our results demonstrate that most naphthalenophanes
bind to AP-sites in DNA and inhibit the APE1-induced hydrolysis of
the latter in vitro. Remarkably, their APE1 inhibitory activity, as
characterized by IC50 and KI values, can be directly related to their
affinity and selectivity to AP-sites, assessed from the fluorescence-
melting experiments. On the other hand, the molecular design of na-
phthalenophanes has crucial influence on their intrinsic AP-site
cleavage activity (i.e., ligand-catalyzed β- and β,δ-elimination
reactions at the AP site), as illustrated by the compounds either
having an exceptionally high AP-site cleavage activity (e.g.,
2,7-BisNP-S, 125-fold more efficacious than spermine) or totally
devoid of this activity (four compounds). Finally, we reveal the unpre-
cedented formation of a stable covalent DNA adduct upon reaction
of one ligand (2,7-BisNP-NH) with its own product of AP-site
cleavage.
proteins leading to generation of DNA–protein crosslinks.[5]
While the latter are thought to be cytotoxic, AP sites are instantly
processed by the base excision DNA repair (BER) pathway,
which eliminates the drug-induced DNA damage and leads to
chemoresistance.[4,6–8] It has been demonstrated that inhibition
of BER increases the sensitivity of cancer cells towards DNA-
alkylating drugs in certain cancers, leading to a therapeutically
useful synergic effect.[9,10] Accordingly, significant efforts have
been devoted to the development of small-molecule inhibitors of
AP endonuclease 1 (APE1, class II AP endonuclease), the key
enzyme of the BER pathway that hydrolytically cleaves DNA at
AP sites leaving a 3′-OH end (Scheme 1, top).[11–13] An alter-
native promising approach to blocking the BER pathway relies
on DNA-binding drugs that selectively bind to AP sites and
thereby mask them, or render them unreactive towards from
APE1. Thus, methoxyamine (MX, or TRC102) and related
alkoxyamines, which form covalent oxime adducts with abasic
sites, demonstrate synergy with DNA-alkylating cytotoxic
drugs,[14,15] and a combination of MX with temozolomide is
currently being evaluated for glioblastoma treatment in a Phase
II clinical trial.[16]
Introduction
Apurinic/apyrimidinic (AP, or abasic) sites represent common
DNA lesions that naturally occur through deglycosylation of DNA,
or through DNA glycosylase-catalyzed removal of damaged
(deaminated, alkylated or oxidized) DNA bases.[1–3] In the
context of chemotherapy employing DNA-alkylating drugs (e.g.,
MMS or temozolomide), the major DNA alkylation product, N7-
methylguanine, undergoes spontaneous depurination leading to
formation of AP sites,[4] as well as a reaction with histone
[a]
C. Caron, X. N. T. Duong, Dr. S. Bombard, Dr. A. Granzhan
CNRS UMR9187, INSERM U1196
Institut Curie, PSL Research University
F-91405, Orsay (France)
E-mail: anton.granzhan@curie.fr
[b]
[c]
C. Caron, X. N. T. Duong, Dr. S. Bombard, Dr. A. Granzhan
CNRS UMR9187, INSERM U1196
Université Paris Sud, Université Paris-Saclay
F-91405, Orsay (France)
Dr. R. Guillot
CNRS UMR8182, Institut de Chimie Moléculaire et des Matériaux
d’Orsay (ICMMO), Université Paris Sud, Université Paris-Saclay
91405 Orsay (France)
Supporting information for this article is given via a link at the end of
the document.
Scheme 1. DNA cleavage at AP sites through APE1-catalysed hydrolysis (top)
and β,δ-elimination catalyzed by secondary amines (bottom).
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
Beyond covalent trapping of abasic sites with reactive drugs
such as MX or analogues, several classes of non-covalent
ligands were shown to selectively bind to abasic sites in DNA
and potentially hinder their recognition and processing by
APE1.[17–22] Along these lines, polyazacyclophanes such as bis-
acridine BisA[23,24] and bis-naphthalenes 2,7-BisNP-O and 2,7-
BisNP-NH (Figure 1) bind to abasic sites with high affinity and
selectivity, outperforming other known ligands.[25,26] We
demonstrated that these compounds inhibit the activity of APE1
in vitro with high efficiency, comparable to that of the best
catalytic APE1 inhibitors.[25] However, in parallel to inhibiting
APE1 activity, polyazacyclophane ligands also induce an
enzyme-independent cleavage of abasic sites via a β-elimination
mechanism, due to the presence of secondary amino groups in
the ligand structure (Scheme 1, bottom).[23,25] The latter process,
reminiscent to the action of AP lyases (class I endonucleases),
was also observed for oligopeptides KWK and KWKK[27–29] as
well as several other small-molecule ligands endowed with
primary or secondary amino groups, eponymously termed
“artificial AP lyases”.[17,30–34] It has been proposed that such
molecules may interfere with the normal BER process because
of the accumulation of “dirty ends” (products of β- and β,δ-
elimination) that cannot be utilized as substrates by DNA
polymerases, and therefore also increase the cytotoxic effect of
DNA-alkylating drugs.[35,36]
series of functionalized macrocycles based on the previously
established 2,7-naphthalenophane scaffold (Figure 1) and
including, among others, ligands conceived either to be devoid
of AP-DNA cleavage activity due to the absence of nucleophilic
groups (2,7-BisNP-O8Me, 2,7-BisNP-O4Gua, 2,7-BisNP-Im) or,
conversely, to possess an enhanced AP-site cleavage activity
(2,7-BisNP-NNH2). Herein, we report the synthesis of novel
ligands, a systematic investigation of their AP-DNA binding
affinity and selectivity, in vitro interference with the APE1
processing of abasic sites, as well as the assessment of their
intrinsic reactivity towards AP-sites in DNA.
Results and Discussion
Synthesis of novel naphthalenophane ligands
All novel naphthalenophanes were prepared from the common
starting material, 2,7-bis(bromomethyl)naphthalene 1 (Scheme
2). The imidazolium-linked cyclophane 2,7-BisNP-Im was
obtained, as a bromide salt, in a two-step reaction of 1 with
imidazole, performed in an analogy to the reported
procedures.[37–40] Kornblum oxidation of 1 gave, in a 52% yield,
naphthalene-2,7-dialdehyde 3, a mutual intermediate for the
synthesis of all polyamine-type macrocycles obtained through a
[2 + 2]-type condensation with the corresponding diamines. Thus,
a reaction of 3 with 2,2′-oxybis(ethylamine), followed by the
reduction of tetraimine intermediate 4 with NaBH4, gave the
previously described macrocycle 2,7-BisNP-O, which was
subsequently converted to the tetramethylated derivative 2,7-
BisNP-O4Me via Eschweiler–Clarke reaction (HCHO/HCOOH)
or to the octamethylated analogue 2,7-BisNP-O8Me by
exhaustive methylation (MeI, DIPEA).
A
tetraguanidine
derivative 2,7-BisNP-O4Gua was obtained, as a hydrochloride
salt, through a treatment of 2,7-BisNP-O with N,N′-di-Boc-S-
methylisothiourea in the presence of HgCl2, followed by the
removal of Boc groups (HCl in CHCl3). Finally, three derivatives
of hexaazamacrocycle 2,7-BisNP-NH with varied substituents at
two middle nitrogen atoms (2,7-BisNP-NMe, -NNH2 and -NOH)
were prepared by condensation of 3 with the correspondingly
substituted derivatives of diethylenetriamine, followed by NaBH4
reduction and conversion to water-soluble hydrochloride salts.
The structures of 2,7-BisNP-Im (2 Br−) and 2,7-BisNP-
O4Gua × 4 HCl were investigated by single-crystal X-ray diffrac-
tion analysis. In the case of the imidazolium-linked cyclophane,
both naphthalene units were found to be strictly coplanar, and
the two anti-oriented imidazolium moieties were nearly perpen-
dicular to the plane of naphthalene units (Figure 2a).
Remarkably, a coplanar geometry of aromatic units is without
precedent in the well-studied series of bis-imidazolium cyclo-
phanes since, to the best of our knowledge, all previously
described anti-conformers adopted a displaced anti-parallel
Figure 1. Structures of previously reported and (*) novel polyazacyclophane
ligands. The counter-ions are omitted for clarity.
1
conformation of aromatic units.[38–40] At the same time, H NMR
spectra of 2,7-BisNP-Im revealed a sharp singlet for methylene
protons at room temperature (δ = 5.77 ppm, in [D6]DMSO),
giving evidence of a rapid interconversion of all possible
conformations (presumably, due to the intramolecular rotation of
imidazolium rings) on the NMR time scale.
In order to fully understand the action of ligands targeting
abasic sites in biological systems, it would be preferable to
decouple the two aforementioned effects, namely the inhibition
of the APE1-induced hydrolysis and the ligand-promoted
cleavage of AP sites. Towards this end, we designed a novel
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
Scheme 2. Synthesis of functionalized naphthalenophane ligands. Reagents and conditions: (i) imidazole, MeOH, reflux, 22 h; (ii) 1, MeCN, 90 °C, 18 h; (iii)
NaHCO3, DMSO, 100 °C; (iv) O(CH2CH2NH2)2, MeCN, room temp., 4 days; (v) NaBH4, DCM–MeOH, room temp.; (vi) HCHO, HCOOH, 120 °C, 48 h; (vii) MeI
(excess), DIPEA, DMF, 60 °C, 24 h; (viii) BocN=C(SMe)NHBoc, HgCl2, NEt3, DCM, room temp., 48 h; (ix) HCl (dioxane), CHCl3, room temp., 24 h; (x)
RN(CH2CH2NH2)2, MeCN, room temp., 4 days; then HCl (dioxane), MeOH.
between the protonated guanidinium groups and oxygen atoms
of the linkers.
a)
b)
AP-DNA affinity of ligands assessed by fluorescence-
melting experiments
The interaction of ligands with abasic sites was studied through
fluorescence-melting experiments with double-stranded oligo-
nucleotides 17-NΦ (Table 1, Φ = THF abasic site), bearing
abasic sites opposite to different orphan residues (N = T, C, A or
G),
a quencher (TAMRA) on the abasic strand, and a
fluorophore (6-FAM) on the complementary strand. The use of
the stable THF analogue of AP sites in these experiments
avoids the ligand-promoted DNA cleavage during the melting
ramps, and a relatively low ionic strength of the buffer (20 mM K+,
pH 7.2) allows the observation of even weak stabilization effects,
expressed as ligand-induced increase of denaturation
temperature (∆Tm). In addition, AP-site selectivity of ligands with
respect to well-matched DNA was estimated through analysis of
the drop of ∆Tm values observed in the presence of unlabeled
double-stranded competitor (calf thymus DNA, 200- or 650-fold
excess with respect to the AP target). In parallel, thermal
denaturation experiments performed with the fully matched
duplex (17-TA) provide another measure of the selectivity of
ligands with respect to undamaged DNA.
Figure 2. Solid-state structures (ORTEP plots, from single crystal X-ray
diffraction analysis): a) 2,7-BisNP-Im (2 Br−), viewed along to the
crystallographic c axis; b) 2,7-BisNP-O4Gua × 4 HCl × 5 H2O, viewed along
the crystallographic b axis. Thermal ellipsoids are drawn at 70 % probability;
cyan lines: intramolecular hydrogen bonds. Hydrogen atoms, counter-ions and
solvent molecules are omitted for the sake of clarity.
The tetraguanidine macrocycle crystallized with naphthalene
units in a displaced anti-parallel conformation (Figure 2b). The
conformation of linkers reveals
a significant potential of
molecular flexibility in aqueous solutions, in spite of the
intramolecular hydrogen bonds observed, in the solid state,
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
TΦ
ticular upon binding to 17-TΦ (∆Tm ≤ 8 °C) and 17-GΦ
Table 1. Sequences and melting temperatures of 17-mer oligonucleotides
used in this work.[a]
(∆TmGΦ ≤ 4 °C). The loss of efficiency was most pronounced
TΦ
GΦ
5′-F–CCAG TTC GNA GTA ACCC-3′
3′-Q-GGTC AAG CYT CAT TGGG-5′
in the case of 2,7-BisNP-O4Me (∆Tm = 5.4 °C, ∆Tm
=
2.0 °C). These observations support the structural model
observed upon binding of 2,7-BisNP-O to a T·T mis-
match[42] and suggest that the substituents at the benzylic
amino groups impede the hydrogen bonding of protonated
amines to carbonyl groups of orphan G or T residues, redu-
cing the overall stabilization of the ligand–AP site complex.
Acronym
N
Y[b]
Tm0 / °C
17-TΦ
17-CΦ
17-GΦ
17-AΦ
17-TA
17-TU
17-TX
T
C
G
A
T
T
T
Φ
Φ
Φ
Φ
A
32.0 ± 0.1[c]
31.9 ± 0.1[c]
33.7 ± 0.1[c]
33.8 ± 0.2[c]
45.0 ± 0.1[c]
n.d.
12
17-TΦ
17-CΦ
17-GΦ
17-AΦ
10
8
6
dU
X
4
46.6 ± 0.2[d]
2
0
[a] In the fluorescence-melting assay, F = 6-FAM, Q = TAMRA; in fluorimetric
DNA cleavage studies, F = ATTO390, Q = DABCYL. [b] Φ = THF abasic site;
X = native abasic site. [c] In 10 mM KAsO2Me2, 10 mM KCl, pH 7.2 (conditions
of the fluorescence-melting assay). [d] In 10 mM KAsO2Me2, 150 mM KCl, pH
7.2 (conditions of the fluorimetric DNA cleavage assay). Data are means ±
s.d. from three technical replicates.
12
10
8
6
4
2
0
The results of fluorescence-melting experiments are
presented in Figure 3. All ligands stabilized the abasic duplexes,
to an extent depending on the base facing the AP site: 17-TΦ >
17-CΦ > 17-GΦ > 17-AΦ. Remarkably, none of the ligands of
the 2,7-BisNP series stabilized the well-matched duplex 17-TA,
in contrast to ethidium bromide (used as a control, non-selective
DNA binder) and 1,5-BisNP-O, which, in spite of its structural
similarity, has a different substitution pattern of naphthalene
units and whose non-specific binding to fully matched DNA was
previously documented.[41] In addition, the selectivity of most
ligands of the 2,7-BisNP series for abasic oligonucleotides was
confirmed by the fact that the presence of double-stranded
competitor only slightly decreased the ligand-induced stabiliza-
tion effect. Considering the effect of substituents introduced into
the naphthalenophane scaffold with respect to parent ligands
(2,7-BisNP-NH and 2,7-BisNP-O), the following observations
emerge from the results presented in Figure 3:
12
10
8
6
4
2
0
10
8
6
4
2
0
8
17-TA
1) Macrocycles bearing substituents at middle nitrogen atoms
of the linkers (2,7-BisNP-NMe, -NNH2, -NOH) demonstrate
the stabilization effect equal to, or even higher than the one
induced by “parent” ligands, for all abasic substrates (e.g.,
6
4
2
TΦ
0
∆Tm ≈ 10 °C), in combination with excellent selectivity
-2
with respect to competitor DNA. This gives evidence that
side chains introduced in the middle of polyamine linkers do
not hamper the binding of naphthalenophanes to AP sites.
In contrast, the sulfur analogue 2,7-BisNP-S was both less
efficient (e.g., ∆TmTΦ ≈ 8 °C) and less selective, with respect
to all abasic substrates.
2
P
sN
m
-S
-O
-O
-I
-NH
4Me 8Me
-O
-NMe
4Gua
-NOH
omide
-NNH
KWKK
-O
-O
2,7-Tri
um br
di
2,7-BisNP
2,7-BisNP
1,5-BisNP
2,7-BisNP
2,7-BisNP
2,7-BisNP
2,7-BisNP
2,7-BisNP
2,7-Bis2N,7P-BisNP
2,7-BisNP
Ethi
Figure 3. Ligand-induced shifts of melting temperature (∆Tm) of abasic oligo-
nucleotides (17-NΦ) or the well-matched control (17-TA), in the absence (dark
bars) or in the presence of ct DNA competitor (medium bars: 20 µM bp, light
bars: 65 µM bp). Conditions: c(oligo) = 0.5 µM, c(ligand) = 1 µM in 10 mM
KAsO2Me2, 10 mM KCl buffer, pH 7.2. Data are means ± s.d. from three
technical replicates.
2) The stabilization induced by the ligands bearing substi-
tuents at the benzylic amino groups (R1, R2 ≠ H, i.e., 2,7-
BisNP-O4Me, 2,7-BisNP-O8Me, 2,7-BisNP-O4Gua) was
lower, compared to the “parent” ligand 2,7-BisNP-O, in par-
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
3) Quite remarkably, the effect of the bicyclophane 2,7-TrisNP
and the bis-imidazolium cyclophane 2,7-BisNP-Im is very
similar to the one of polyazacyclophane ligands, in spite of
their significantly different shapes. In the case of 2,7-
TrisNP, this may be rationalized taking into account its
shape similarity to cylindrical metallohelicates which bind to
abasic sites with high affinity, even though the structural
details of their binding mode have not yet been
elucidated.[22] As for 2,7-BisNP-Im, this ligand may bind to
AP sites in a conformation which is significantly different
from the one observed in the solid state (cf. Figure 2a),
most likely in a folded form allowing the insertion of one of
naphthalene units inside the abasic pocket.
a)
5′
3′
5′
5′
Φ
Q
Φ
Q
F
APE1
+
17-TΦ
KI
3′
F
ligand
5′
3′
Q
F
Φ =
1.0
0.8
0.6
0.4
0.2
0.0
b)
2,7-BisNP-O8Me
IC50 = 3.9 µM
Inhibition of enzymatic DNA cleavage
To study the competitive effect of the ligands on the enzymatic
processing of AP sites, we used a real-time fluorimetric assay
similar to the one employed in our previous work.[25] In this assay,
APE1-induced cleavage of the abasic substrate 17-TΦ (Table 1)
is monitored through the fluorescence increase which occurs
upon cleavage and dehybridization of the quencher-bearing
strand (Figure 4a). However, in preliminary experiments we
observed that oligonucleotides labelled with 6-FAM, as well as
with several other conventional fluorophores such as Alexa Fluor
(AF) 488, AF 546, AF 594, TAMRA, or Texas Red, underwent
significant quenching upon addition of most ligands. While this
ligand-induced quenching is tolerable in fluorescence-melting
experiments (which aim at measuring the temperature of the
melting transition and do not account for the exact fluorescence
intensity), this phenomenon could bias the output of kinetic
assays. Therefore, in order to minimize the extent of ligand-
induced quenching, ATTO390 was employed as the fluorophore
least prone to this phenomenon, in combination with DABCYL
quencher.[43]
2,7-BisNP-O
IC50 = 0.18 µM
0.1
1
10
Ligand concentration (µM)
Figure 4. Inhibition of APE1 activity by AP-site ligands. a) Principle of the real-
time fluorimetric assay for APE1 activity and indirect inhibition of APE1 by
ligands binding to AP-sites. F = ATTO390, Q = DABCYL, Φ = THF abasic site.
b) Representative dose-response plots for APE1 inhibition by 2,7-BisNP-O
(black) and 2,7-BisNP-O8Me (red). Conditions: c(17-TΦ) = 0.2 µM in 50 mM
HEPES, 50 mM NaCl, 1 mM MgCl2, 2 mM DTT buffer, pH 8.0, T = 37 °C,
λex = 395 nm, λem = 465 nm.
Table 2. Inhibitory activity of macrocyclic ligands with respect to APE1-
induced cleavage of 17-TΦ.[a]
The results of the kinetic assay demonstrated that all ligands
were able to inhibit APE1 activity in a dose-dependent manner
(Figure 4b and Supporting Information, Figure S1), with the
corresponding IC50 values ranging from 0.14 µM (2,7-BisNP-
NNH2) to 89 µM (2,7-BisNP-O4Me, Table 2). Taking into
account the fact that indirect inhibition is described by the same
mathematical model as competitive inhibition (i.e., Vmax = const,
Ligand
IC50 / µM
0.19
0.18
55
KI / nM
52
2,7-BisNP-NH
2,7-BisNP-O
50
2,7-BisNP-S
15000
52
app
KM = αKM, where α = 1 + [I] / KI), with the only difference that
2,7-BisNP-NMe
2,7-BisNP-NNH2
2,7-BisNP-NOH
2,7-BisNP-O4Me
2,7-BisNP-O8Me
2,7-BisNP-O4Gua
2,7-BisNP-Im
1,5-BisNP-O
0.19
0.14
0.46
89
the inhibition constant (KI) represents the dissociation constant
of the inhibitor–substrate (and not inhibitor–enzyme) complex,[25]
we calculated the corresponding KI values (Table 2) using the
well-known relationship (Eq. 1):[44]
39
130
25000
1100
110
550
720
77
IC50 = KI (1 + S / KM),
(1)
3.9
where S is concentration of the substrate (0.2 µM) and KM is
Michaelis constant for APE1 (KM = 76 nM, Supporting Infor-
mation, Figure S2).[45] Of note, the KI values for the macrocycles
2,7-BisNP-NH and 2,7-BisNP-O (≈50 nM) were of the same
order as the ones obtained in our previous work (34 nM) despite
a significantly lower concentration of the substrate (25 nM).[25]
0.4
2.0
2.6
2,7-TrisNP
0.28
[a] Conditions: see Figure 4 caption.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
Remarkably, a good correlation could be observed between
the 1/KI values (i.e., affinity constants) and the ligand-induced
stabilization of the substrate (∆TmTΦ) observed in the presence of
excess competitor (r 2 = 0.72, Figure 5a): in other words, potent
and selective AP-site ligands, whose stabilizing effect was not
depressed by the presence of the competitor (2,7-BisNP-NNH2,
2,7-BisNP-NH, 2,7-BisNP-O, 2,7-BisNP-NMe), were most
efficient at inhibiting the APE1-induced cleavage of AP-sites. A
correlation made using ∆TmTΦ values obtained in the absence of
the competitor was unsatisfactory (r 2 = 0.26, Figure 5b) due to
the fact that the stabilization induced by non-selective binders
(e.g., 1,5-BisNP-O) was strongly reduced in the presence of the
competitor. Altogether, the comparison made with a series of
different ligands, varying in terms of their affinity and selectivity
to AP-sites, convincingly support the indirect inhibition model
and highlight the efficiency of DNA-targeted inhibition of DNA
repair. Moreover, the revealed interplay between the affinity and
the selectivity of DNA-targeting inhibitors is of utmost importance
for their application in the cellular context, where DNA damage
sites are disseminated throughout the large excess of
undamaged genome.
5 µM (QF, Table 2). This factor was taken into account for
calculation of corrected initial reaction rates (vcorr, Table 3), used
as quantitative indicators of AP-DNA cleavage activity of ligands
(vcorr = v0 / QF, where v0 are the initial slopes of real-time
fluorescence intensity plots, Figure 6b).
a)
5′
Q
+
+
5′
3′
X
Q
F
ligand
5′
5′
(β)
or
17-TX
(δ)
X =
3′
F
1000
800
600
400
b)
2,7-BisNP-S
2,7-BisNP-NOH
2,7-BisNP-NH
a)
b)
12
10
8
12
10
8
2,7-BisNP-O4Gua
r 2 = 0.72
r 2 = 0.26
none
6
6
4
4
0
50
100
150
2
2
Time (min)
0
0
0.00
0.01
1 / KI (nM-1
0.02
0.03
0.00
0.01
1 / KI (nM-1
0.02
0.03
Figure 6. Real-time fluorimetric assay for determination of AP-DNA cleavage
activity of ligands. a) Principle of the assay (F = ATTO390, Q = DABCYL, X =
native abasic site). b) Representative fluorimetric readout for cleavage of 17-
TX with 2,7-BisNP-NH, 2,7-BisNP-S, 2,7-BisNP-NOH and 2,7-BisNP-O4Gua.
Conditions: c(17-TX) = 0.2 µM, c(ligand) = 5 µM in 10 mM KAsO2Me2, 150 mM
KCl buffer, pH 7.2, T = 37 °C, λex = 395 nm, λem = 465 nm; the ligand was
added at t = 0. The transient drop of fluorescence intensity after ligand
addition (red curve) is due to the ligand-induced quenching effect. The scale of
fluorescence intensity (max value = 1000 a.u.) was adjusted in order to allow
the best comparison of initial reaction rates induced by different ligands.
)
)
Figure 5. Correlation of ligand affinity constants to AP-sites, derived from
APE1-inhibition studies interpreted in terms of indirect inhibition model (Ka
1 / KI) with ligand-induced stabilization of 17-TΦ observed a) in the absence
=
and b) in the presence of competitor (ct DNA, 65 µM bp).
Intrinsic AP-DNA cleavage activity of ligands
The intrinsic activity of ligands with respect to nucleophile-
catalyzed cleavage of AP-sites was also assessed by means of
a real-time fluorimetric assay, similar to the one developed by
Berthet et al.[46] Our variant of this assay employs the 17-mer
duplex 17-TX (Table 1), double-labelled with ATTO390 and
DABCYL and bearing a native abasic site (X), quantitatively
produced by UDG treatment of the corresponding uracil-
containing precursor 17-TU (Supporting Information, Figure S3).
As in the APE1 activity assay, the progress of ligand-induced
DNA cleavage is monitored through an increase of fluorescence,
taking place upon dehybridization of the cleaved strand (Figure
6a). In order to compensate for the partial ligand-induced fluoro-
phore quenching, the extent of this effect was measured in
control experiments with a non-cleavable substrate (17-TΦ) and
quantified as a ratio of fluorescence intensity in the presence
and in the absence of ligands, used at identical concentration of
In parallel, we assessed the DNA-cleavage activity of ligands
through PAGE analysis using the duplex 17-TX, in which the AP
strand was 5′-32P-labelled. This complementary assay,
performed in the conditions similar to the ones of the fluorimetric
assay (with a fixed incubation time of 1 h), avoids the problems
related to the use of fluorescently labeled DNA (ligand-induced
quenching) and, in addition, allows to estimate the distribution of
various products of strand cleavage (β- and β,δ-elimination
products) and secondary reactions. Based on the results of the
real-time fluorimetric assay (Table 3) and the PAGE assay
(Figure 7), all ligands could be divided into three groups with
respect to their AP-site cleavage activity:
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
showed a discrepancy between the two methods: strong
cleavage activity was detected by the PAGE assay (con-
version = 89% after 1 h), but not in the fluorimetric assay,
presumably due to its strong quenching effect (QF = 0.28).
Table 3. DNA AP-site cleavage activity of macrocycles and reference ligands,
measured by the real-time fluorimetric assay.[a]
Compound
v0 / a.u. s−1 [b]
QF [c]
vcorr / a.u. s−1 [d]
2,7-BisNP-NH
2,7-BisNP-O
2,7-BisNP-S
3.02 ± 0.16
5.64 ± 0.34
115.92 ± 5.87
4.29 ± 0.44
33.00 ± 1.23
12.85 ± 0.65
1.08 ± 0.08
0.16 ± 0.02
0.31 ± 0.19
0.14 ± 0.05
8.57 ± 0.15
0.09 ± 0.01
0.17 ± 0.02
1.31 ± 0.11
0.10 ± 0.00
0.97
0.89
0.75
0.90
0.95
0.78
0.91
0.79
0.85
0.54
0.73
0.28
0.97
1.04
3.13
6.35
155
a)
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17
*
**
2,7-BisNP-NMe
2,7-BisNP-NNH2
2,7-BisNP-NOH
2,7-BisNP-O4Me
2,7-BisNP-O8Me
2,7-BisNP-O4Gua
2,7-BisNP-Im
1,5-BisNP-O
2,7-TrisNP
4.76
34.9
16.5
1.19
0.20
0.36
0.25
11.8
0.31
0.17
1.25
0.10
β
δ
†
b)
100
80
60
40
20
0
β
†
**
*
δ
e
e
a
P
2
H
H
M)
KK ine
perm
Me
OH
isN
P-S
P-O
isN
P-O
N
sN
P-Im
KW
P-N
ontrol
0.5
c
27 nM)
P-N
isN
P-N
isN
isN
(
P-N
P-O4MP-O8M
s
P-O4Gu
isN
isN
2,7-Tri
isN
untd
isN isN
aOH
2,7-B
1,5-B
2,7-B
2,7-B
isN
2,7-B
KWKK
2,7-B
N
APE1 (0.
2,7-B
2,7-B
2,7-B2,7-B
2,7-B
spermine × 4 HCl
no ligand
Figure 7. AP-site cleavage activity of ligands assessed by PAGE. a) PAGE
analysis of 32P-labelled 17-TX following incubation with ligands. Lane
assignment: 1) untreated control; 2) NaOH treatment (0.5 M); 3) APE1 (0.27
nM); 4) 2,7-BisNP-O; 5) 2,7-BisNP-NH; 6) 2,7-BisNP-S; 7) 1,5-BisNP-O; 8)
2,7-TrisNP; 9) 2,7-BisNP-NMe; 10) 2,7-BisNP-O4Me; 11) 2,7-BisNP-O8Me;
12) 2,7-BisNP-O4Gua; 13) 2,7-BisNP-NNH2; 14) 2,7-BisNP-NOH; 15) 2,7-
BisNP-Im; 16) KWKK; 17) spermine. Conditions: c(17-TX) = 0.2 µM, c(ligand)
= 5 µM, incubation for 1 h in 10 mM KAsO2Me2, 150 mM KCl buffer, pH 7.2,
T = 37 °C. Left and right panels are parts of the same gel. b) Quantification of
PAGE data (means ± s.d. from three technical replicates). Band assignment /
color codes: * / grey, uncleaved substrate; β / blue, product of β-elimination; δ
/ cyan, product of β,δ-elimination; † / green, 3′-OH product of APE1 cleavage;
** / pink, putative ligand cross-link.
[a] Substrate: 17-TX (F = ATTO390. Q = DABCYL), conditions: see Figure 6
legend. [b] Initial slope of the real-time fluorescence intensity plot (mean ± s.d.
from three technical replicates). [c] Ligand-induced fluorescence quenching of
substrate, QF = Fwith ligand / Fno ligand, measured with the non-cleavable analogue
17-TΦ at c(ligand) = 5 µM. [d] Quenching-adjusted initial rate of AP-DNA
cleavage reaction.
1) Very efficient AP-cleaving ligands, namely 2,7-BisNP-S
(vcorr = 155 a.u. s−1) and 2,7-BisNP-NNH2 (vcorr = 34.9 a.u.
s−1), for which a total conversion of the substrate was
observed after 1 hour, with the products of β- and β,δ-
elimination forming in approximately 2 : 1 ratio (according to
PAGE analysis). The high efficiency of 2,7-BisNP-NNH2 is,
expectedly, due to the presence of two additional primary
amino groups, which lead to an 11-fold higher cleavage rate
with respect to the parent ligand 2,7-BisNP-NH (vcorr = 3.11
a.u. s−1). The exceptionally high activity of the sulfur-
containing macrocycle is particularly remarkable, taking into
account the fact that its affinity for AP sites is significantly
lower than the one of 2,7-BisNP-NH and 2,7-BisNP-NNH2,
as demonstrated by both fluorescence-melting experiments
(Figure 3) and APE1 inhibition studies (Table 2).
3) Mostly inactive ligands (vcorr < 2 a.u. s−1, < 10% substrate
conversion in PAGE assay): 2,7-BisNP-O4Me, 2,7-BisNP-
O8Me, 2,7-BisNP-O4Gua, 2,7-BisNP-Im, spermine and
KWKK. Although spermine and the tetrapeptide KWKK are
well-established AP-DNA cleaving agents, in both of our
assays their activity was only marginal, likely due to at least
100-fold lower concentration, comparing with previous
works.[29,47] This is also in line with the negligible AP-site
binding effect of KWKK observed in FRET-melting
experiments (Figure 3).
These results clearly demonstrate the influence of molecular
structure of naphthalenophanes on their AP-DNA cleavage
activity. Specifically, introduction of side chains at central
nitrogen atoms of linkers results in enhanced cleavage activity
(2,7-BisNP-NMe, 2,7-BisNP-NOH and, particularly, 2,7-BisNP-
NNH2) compared to 2,7-BisNP-NH. Conversely, suppression of
nucleophilic centers through introduction of substituents at
benzylic amino groups (2,7-BisNP-O8Me, 2,7-BisNP-O4Gua) or
2) Moderately efficient AP-cleaving ligands (2
< vcorr <
20 a.u. s−1): 2,7-BisNP-NH, 2,7-BisNP-NMe, 2,7-BisNP-
NOH, 2,7-BisNP-O and 1,5-BisNP-O, which induced from
~60 to ~90% conversion of the substrate after 1 h (as
determined by PAGE), giving predominantly β-elimination
products (β : δ = 4.5 to 8). Only one ligand, 2,7-TrisNP,
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
removal of these groups (2,7-BisNP-Im) totally abolishes the
AP-DNA cleavage activity. In the last series, the derivative 2,7-
BisNP-O4Me represents an intermediate case: despite the
presence of four tertiary, putatively nucleophilic, amino groups,
the AP-DNA cleavage activity of this ligand is strongly reduced
with respect to 2,7-BisNP-NH (vcorr = 1.19 a.u. s−1, conversion =
6%). This behavior is likely due to the combination of two
phenomena, namely: (i) the tertiary amino groups do not allow
the formation of the pre-incision iminium intermediate of the β-
elimination reaction (cf. Scheme 1); (ii) the AP-DNA affinity of
2,7-BisNP-O4Me is reduced with respect to the other ligands of
the series (cf. Figure 3 and Table 2). Notably, a similar decrease
of affinity and DNA-cleaving activity upon mono-methylation of
secondary amino groups was observed in the series of
nucleobase–polyamine–intercalator conjugates.[17]
to 18% after 5 h of incubation), giving evidence that the β-
elimination product represents an intermediate for the formation
of the covalent adduct (and not vice versa).
a)
1
2
3
4
5
6
7
***
*
**
β
δ
b)
1
2
3
4
5
6
7
8
9
10 11
Formation of covalent adducts with abasic sites
The results of PAGE analysis revealed that, in some cases, the
formation of expected products of AP-site cleavage (β- and β,δ-
elimination products) was accompanied by appearance of novel
bands of intermediate mobility (** in Figure 7). These bands
were observed mostly with 2,7-BisNP-NH (20% yield, lane 5),
but trace amounts could also be detected with 2,7-BisNP-S
(5%) and 2,7-BisNP-O (< 1%). In order to get insight into the
nature of these products, we performed several complementary
studies. Incubation of 2,7-BisNP-NH with a DNA substrate
devoid of abasic site (17-TU) gave no evidence of strand
cleavage (Figure 8a, lanes 6–7), therefore excluding the
possibility of a strand cleavage at a secondary site (i.e., away
from the AP site). Furthermore, the product was stable towards
a brief heating at 90 °C (Figure 8a, lane 4), which speaks in
favor of a covalent modification of DNA. Next, we performed a
trapping reaction through incubation of 17-TX with 2,7-BisNP-
NH in the conditions of reductive amination (NaCNBH3, followed
by NaBH4 quenching; Figure 8a, lane 5).[29,33] In this case, we
evidenced the formation of a slower-migrating band (*** in
Figure 8a), supposedly corresponding to the reduced pre-
incision iminium base intermediate (cf. Scheme 1), as well as
another band which migrated closely to the ** band and which
could be attributed to the reduced post-incision intermediate (i.e.,
the product of reductive amination reaction of the polyamine with
the α,β-unsaturated aldehyde).[29,48] Taking into account these
results, we may suggest that the ** band in lanes 3–4 (Figure
8a), formed in the absence of reducing agents, corresponds to a
stable but unreduced covalent adduct of the ligand with its own
product of AP-site cleavage via β-elimination. In addition, its
greatly retarded migration with respect to the other cleavage
products is consistent with the partial charge neutralization of
the oligonucleotide imparted by the remnant of the polycationic
ligand (3 to 4 positive charges).
*
**
β
δ
c)
100
80
60
40
20
0
**
δ
β
*
0
1
2
3
4
5
16
Time (h)
Figure 8. Formation of a covalent adduct of 2,7-BisNP-NH with abasic sites.
a) PAGE analysis showing formation of slow-migrating bands of 2,7-BisNP-
NH upon incubation with 17-TX. Lane assignment: 1) 17-TX incubated with 0.5
M NaOH (1 h); 2) untreated 17-TX; 3) 17-TX incubated with 2,7-BisNP-NH
(5 µM); 4) same, but heated at 90 °C (5 min) prior to PAGE analysis, 5) 17-TX
incubated with 2,7-BisNP-NH (5 µM) and NaCNBH3 (25 mM), followed by
addition of NaBH4 (100 mM) prior to PAGE analysis; 6) untreated 17-TU; 7)
17-TU incubated in the presence of 2,7-BisNP-NH (5 µM). b) Time
dependence of the formation of covalent adduct. Lane assignment: 1) 17-TX
incubated without additives for 16 h; 2) 17-TX incubated with 0.05 M NaOH
(1 h); 3–11) 17-TX incubated with 2,7-BisNP-NH for ¼, ½, ¾, 1, 2, 3, 4, 5 and
16 h, respectively. In all experiments, c(17-TX) = 0.2 µM. Band assignment: *,
unmodified substrate; **, covalent ligand–DNA adduct; ***, reduced imine
intermediate (lane 5), β, product of β-elimination (or its reduced form in lane
5); δ, product of β,δ-elimination. c) Quantification of PAGE shown in panel b).
In a separate experiment, we examined the influence of
reaction time on the formation of this product. The adduct yield
increased with time and reached its maximum (~60%) after 5 h
of incubation at 37 °C, when all substrate was consumed (Figure
8b,c). Notably, the increase of the amount of the DNA–ligand
adduct was accompanied by a decrease of the yield of the β-
elimination product at longer incubation time (from 30% after 1 h
Finally, we attempted mass-spectrometric characterization of
the product corresponding to the ** band observed in PAGE. A
crude reaction mixture of 17-TX with 2,7-BisNP-NH was
concentrated, desalted (EtOH precipitation followed by reverse-
solid-phase extraction), and analyzed by LC/MS (Figure 9a).
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
Along with an unresolved peak containing cleavage products
and the complementary strand, we observed a minor peak with
a higher retention time (tR = 4.05 min), consistent with the more
lipophilic character imparted by the presence of the naph-
thalenophane remnant. A corresponding mass spectrum (Figure
9b) revealed the presence of two molecular ion peaks (m/z =
1036 and 1555), which could indeed correspond to adduct of the
ligand with the β-elimination product of DNA cleavage (M =
3111 Da). Although the structure of this product could not be
established unambiguously at this stage, we may suggest that it
could be formed through a 1,4-conjugate addition of the neigh-
boring amino group of the macrocyclic ligand to the post-incision
iminium intermediate (cf. Scheme 1) leading to formation of a
cyclic 1,4-diazepine (Figure 9c), which may be subjected to
further isomerization. This hypothesis explains the fact that,
among all ligands, only 2,7-BisNP-NH (presenting a 1,2-diamino
fragment) gives substantial yield of the covalent adduct.
formation of the cross-link was proposed to occur through a 1,4-
conjugate addition of N1 of a proximate adenine residue to the
unsaturated aldehyde product of β-elimination cleavage, which
is reminiscent to the mechanism proposed above.
Conclusions
In the present work, we systematically explored the 2,7-naphtha-
lenophane scaffold as a versatile platform for the development
of functionalized ligands interacting, in various ways, with AP-
sites in DNA. We demonstrated that careful molecular
engineering could be exploited to exacerbate or alleviate some
of their biochemical properties, as summarized in Table 4.
Specifically, among seven novel compounds based on this
scaffold, most ligands conserved the high affinity to abasic sites
and excellent selectivity with respect to well-matched DNA.
Furthermore, we demonstrated that most compounds were able
to inhibit the APE1-induced cleavage of AP-sites in DNA by
competing with the enzyme for binding to the substrate, and that
the inhibitory activity (as characterized by IC50 and KI values) is
directly related to the affinity and selectivity of ligands to the
substrate (i.e., abasic site). Beyond direct implications in the
context of APE1-related chemoresistance, these conclusions
may be extended to other DNA repair pathways whose DNA
substrate may represent a target for small-molecule ligands,
such as direct damage reversal[53] and mismatch repair.[54,55]
3.36
a)
3.47
unmodified DNA
%
**
0.72
4.05
0.00
100
1.00
2.00
3.00
Time (min)
4.00
5.00
6.00
1114.4
b)
[M−3H]3–
1036.3
[M−2H]2–
1555.5
1554.8
%
1037.1
Table 4. Qualitative summary of AP-DNA-binding properties, APE1 inhibitory
activity, and intrinsic reactivity of naphthalenophane macrocycles with respect
to AP-sites in DNA.
1555.8
1556.3
1600
Ligand
AP-DNA binding
APE1
inhibi-
tion[b]
AP-DNA
cleavage
activity[c]
AP
cross-
linking
0
1000
1100
1200
1300
1400
1500
1700
1800
1900
affinity[a]
+++
+++
++
select.
m/z
c)
2,7-BisNP-NH
2,7-BisNP-O
++
+++
+
+++
+++
+
+
++
[d]
+
±
M
= 3111.4
[d]
2,7-BisNP-S
+++
+
±
2,7-BisNP-NMe
2,7-BisNP-NNH2
2,7-BisNP-NOH
2,7-BisNP-O4Me
2,7-BisNP-O8Me
2,7-BisNP-O4Gua
2,7-BisNP-Im
1,5-BisNP-O
+++
+++
+++
+
++
+++
++
+
+++
+++
+++
+
−
−
−
−
−
−
−
−
−
Figure 9. a) TIC chromatogram of the crude reaction mixture of 17-TX
(0.2 µM) with 2,7-BisNP-NH (50 µM) (see Experimental Section for details). b)
Mass spectrum of the peak with tR = 4.05 min (**). The peak with m/z = 1114.4
could not be assigned. c) Putative structure of the covalent adduct of 2,7-
BisNP-NH to the product of β,δ-elimination.
++
+
−
++
++
++
++
+
++
−
While the reaction of β-elimination products of AP-site
cleavage with external alkoxyamines is known and has been
exploited for their quantitative detection,[49] the formation of
covalent adducts with 2,7-BisNP-NH represents, to the best of
our knowledge, the first example of generation of stable adducts
through reaction of cleaved AP-sites with polyamine ligands. In
this context, it is interesting to mention the formation of stable
interstrand cross-links through reaction of both uncleaved[50,51]
and cleaved[52] AP-sites with purine residues observed in native
(non-reducing) conditions. Notably, in the latter case, the
++
+++
++
−
+
−
+++
++
++
+
2,7-TrisNP
+++
+++
+
[a] ∆TmTΦ < 6 (+), 6 to 9 (++), > 9 °C (+++). [b] KI > 2000 (+), 200 to 2000 (++),
< 200 nM (+++). [c] As discussed in manuscript text. [d] Traces.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
pressure and aq. K2CO3 (6%, 100 mL) was added. The yellow precipitate
was collected and flash-chromatographed using gradient elution (SiO2;
eluent: 0 to 10% MeOH in CH2Cl2), to give the product as a yellow
powder (640 mg, 2.20 mmol, 44%), m.p. 139–146 °C. 1H NMR (300 MHz,
CDCl3): δ 5.28 (s, 2H), 6.94 (s, 1H), 7.12 (s, 1H), 7.29 (d, 1H), 7.53 (s,
1H), 7.60 (s, 1H), 7.83 (d, 1H); 13C NMR (75 MHz, CDCl3): δ 50.9 (CH2),
125.5 (CH), 126.2 (CH), 128.8 (CH), 130.1 (CH), 132.6 (Cq), 133.2 (Cq),
134.7 (Cq); MS (ESI+): m/z (%) 289.2 (100) [M + H]+.
We also demonstrated that deliberate incorporation of
substituents into the naphthalenophane scaffold may be
exploited in order to modulate, or totally suppress, the intrinsic
AP-site cleavage activity of ligands (β- and β,δ-elimination
reactions at AP-sites). Quite unexpectedly, most efficient AP-site
cleavage was observed with the unsubstituted dithiaazamacro-
cycle 2,7-BisNP-S, followed by the rationally designed ligand
endowed with two auxiliary nucleophilic groups, 2,7-BisNP-
NNH2. Even though the reasons of the particularly high activity
of 2,7-BisNP-S are not clear at this point, this observation
prompts for a further systematic and mechanistic study of sulfur-
containing DNA ligands as AP site-cleaving agents.
Last but not least, we demonstrated the formation of an
unexpected covalent adduct upon reaction of at least one ligand,
2,7-BisNP-NH, with its own product of β-elimination cleavage at
AP-sites. In analogy to other intrastrand and small-molecule
cross-links of AP-sites, we expect this adduct to be genotoxic,
representing a hurdle for DNA polymerases.[50,52,56] It may also
be potentially harnessed for covalent targeting of abasic sites
and incorporation of reporter moieties, such as fluorophores or
affinity tags for DNA pull-down studies.
1,5(1,3)Diimidazolia-3,7(2,7)dinaphthalenacyclooctaphane
dibro-
mide (2,7-BisNP-Im, 2 Br−): To a solution of compound 2 (157 mg, 0.54
mmol) in acetonitrile (50 mL), 2,7-bis(bromomethyl)naphthalene 1 (171
mg, 0.54 mmol) was added. The resulting mixture was heated under
reflux for 18 h and a white precipitate appeared. After cooling to room
temperature, the precipitate was collected, washed with acetonitrile and
ether, and recrystallized from MeOH–H2O to give the product as a white
solid (87.1 mg, 0.14 mmol, 27%). 1H NMR (300 MHz, [D6]DMSO): δ 5.76
(s, 4H), 7.34 (s, 2H), 7.66 (d, J = 8.4 Hz, 2H), 7.99 (s, 2H), 8.05 (d, J =
8.5 Hz, 2H), 9.53 (s, 1H); 13C NMR (75 MHz, [D6]DMSO): δ 51.8 (CH2),
123.8 (CH), 124.1 (CH), 126.0 (CH), 128.5 (CH), 132.1 (Cq), 133.0 (Cq),
134.6 (Cq), 137.8 (CH); MS (ESI+): m/z (%): 441.0 (12) [M – H]+, 221.2
(100) [M]2+; anal. calcd. for C30H26Br2N4 × 1.25 H2O (624.88): C, 57.66; H,
4.6; N, 8.97; found: C, 57.63; H, 4.39; N, 8.75.
Native AP sites are relatively stable in naked DNA, with a
half-life of more than 3 weeks at 37 °C. However, AP sites
embedded into chromatin undergo an up to 100-fold accelerated
cleavage via β-elimination mechanism, catalyzed by lysine-rich
N-terminal tails of histone proteins, as well as formation of
persistent DNA-protein cross-links.[57–59] In this context, masking
AP sites with non-nucleophilic ligands (e.g., 2,7-BisNP-O4Gua
and 2,7-BisNP-O8Me) may protect them not only from the
APE1-induced hydrolysis, but also from β-elimination cleavage
by histones and DNA lyases. Further studies will shed light on
these aspects of AP reactivity.
Naphthalene-2,7-dicarbaldehyde (3):
A solution of 2,7-bis(bromo-
methyl)naphthalene 1 (14.0 g, 44.5 mmol) and NaHCO3 (37.9 g, 451
mmol) in anhydrous DMSO (125 mL) was heated at 100 °C during 3.5 h
under argon atmosphere. The resulting mixture was cooled to room
temperature, and water (200 mL) was added. The aqueous layer was
extracted with ethyl acetate (3 × 230 mL). The organic phases were
combined, washed with water, dried with MgSO4, and the solvent was
removed under reduced pressure. The residue was purified by flash
chromatography (SiO2; eluent: using 0 to 20 % AcOEt in cyclohexane), to
give the product as a light-yellow powder (4.63 g, 25.1 mmol, 57%), m.p.
142–143 °C (lit.[41] 145–146 °C). 1H NMR (300 MHz, CDCl3): δ 8.03 (d, J
= 8.4 Hz, 2H), 8.13 (dd, J = 8.6 Hz, 2H), 8.52 (s, 2H), 10.22 (s, 2H); MS
(ESI+): m/z (%) 185.1 (100) [M + H]+.
Experimental Section
General procedure for the synthesis of macrocyclic tetraimine
intermediates 4 and 5a–c: A solution of the corresponding diamine (1.0
mmol) in MeCN (60 mL; in the case of a poorly soluble N,N-bis(2-amino-
ethyl)aminopropan-1-ol (7), 4 mL of MeOH were used as a co-solvent)
was added at room temperature dropwise, within 3 h, to a vigorously
stirred solution of 3 (1.0 mmol) in MeCN (60 mL). The reaction mixture
was stirred at room temperature for 5–7 days. The formed precipitate
was collected, washed with MeCN and dried in vacuo, to give the
corresponding macrocyclic tetraimine 5a–c.
General remarks: All commercially available chemicals were reagent
grade and used without further purification. NMR spectra were measured
with a Bruker Avance 300 (1H: 300 MHz, 13C: 75 MHz) spectrometer at
25 °C; chemical shifts are given in ppm (δ) values. Multiplicities of 13C
NMR signals were determined from DEPT-135 experiments. The melting
points were determined in open-end capillaries with a digital melting point
instrument (SMP30, Stuart). Elemental microanalysis of all novel
compounds was performed by the Service de Microanalyse, CNRS–
ICSN, Gif-sur-Yvette, France. Mass spectra (ESI in the positive-ion
mode) were recorded with a Waters ZQ instrument. The synthesis of
naphthalenophanes 2,7-BisNP-NH × 6 HCl, 2,7-BisNP-O × 4 HCl, 2,7-
BisNP-S × 4 HCl, 1,5-BisNP-O × 4 HCl[41] and 2,7-TrisNP × 6 HCl[60]
was described elsewhere. Tetrapeptide KWKK (hydrochloride salt) was
purchased from Eurogentec. Ethidium bromide and spermine tetrahydro-
chloride were purchased from Sigma–Aldrich. Stock solution of ligands
were prepared in water at a concentration of 2 mM and stored at 4 °C.
6,16-Dioxa-3,9,13,19-tetraaza-1,11(2,7)-dinaphthalenacycloicosa-
phane-2,9,12,19-tetraene (4) was obtained in a 91% yield through
reaction of 3 with 1,5-diamino-3-oxapentane. White powder, 1H NMR
(300 MHz, CDCl3): δ 3.78–3.88 (m, 4H), 7.01 (s, 1H), 7.68 (d, J = 8.5 Hz,
1H), 7.98 (d, J = 8.5 Hz, 1H), 8.10 (s, 1H).
6,16-Dimethyl-3,6,9,13,16,19-hexaaza-1,11(2,7)-dinaphthalenacyclo-
icosaphane-2,9,12,19-tetraene (5a) was obtained in
a 92% yield
through reaction of 3 with 2,2′-diamino-N-methyldiethylamine. White
1
Synthesis
powder, H NMR (300 MHZ, CDCl3): δ 2.34 (s, 3H), 2.82 (t, 4H), 3.71 (t,
4H), 6.82 (s, 2H), 7.84 (d, J = 8.6 Hz, 2H), 8.09 (s, 2H), 8.16 (d, J = 8.6
Hz, 2H).
2,7-Bis((imidazol-1-yl)methyl)naphthalene (2): To a solution of 2,7-
bis(bromomethyl)naphthalene 1 (1.57 g, 5 mmol) in MeOH (100 mL),
imidazole (3.4 g, 50 mmol) was added. The resulting mixture was heated
under reflux for 22 h. The solvent was evaporated under reduced
6,16-Bis(3-(tert-butoxycarbonyl)aminopropyl)-3,6,9,13,16,19-hexa-
aza-1,11(2,7)-dinaphthalenacycloicosaphane-2,9,12, 19-tetraene (5b)
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
was obtained in a 54% yield through reaction of 3 with tert-butyl N-{3-
[bis(2-aminoethyl)amino]propyl}carbamate.[61] White powder, 1H NMR
(300 MHz, CDCl3): δ 1.35 (s, 18H), 1.55–1.57 (m, 4H), 2.48 (t, 4H), 2.84–
2.87 (m, 8H), 3.04–3.06 (m, 4H), 3.67 (br s, 8H), 4.76 (br s, NH), 6.78 (s,
4H), 7.87 (d, J = 8.6 Hz, 4H), 8.07 (s, 4H), 8.17 (d, J = 8.6 Hz, 4H).
J = 8.5 Hz, 2H), 7.96–7.99 (s + d, 4H); 13C NMR (75 MHz, D2O): δ 22.0
(CH2), 37.0 (CH2), 42.5 (CH2), 48.2 (CH2), 50.1 (CH2), 50.7 (CH2), 127.9
(CH), 128.4 (Cq), 129.2 (CH), 130.2 (CH), 132.4 (Cq), 133.3 (Cq); MS
(ESI+): m/z (%) 625.7 (65) [M + H]+; anal. calcd. for C38H64Cl8N8 × 4 H2O
(988.65): C, 46.17; H, 7.34; N, 11.33; found : C, 46.47; H, 7.15; N, 10.91.
6,16-Bis(3-hydroxypropyl)-3,6,9,13,16,19-hexaaza-1,11(2,7)-dinaph-
thalenacycloicosaphane-2,9,12,19-tetraene (5c) was obtained in a
52% yield through reaction of 3 with N,N-bis(2-aminoethyl)aminopropan-
1-ol 7. Light-yellow powder, H NMR (300 MHz, CDCl3): δ 1.75–1.78 (m,
4H), 2.75-2.78 (m, 4H), 2.89-2.91 (m, 8H), 3.73 (br s, 8H), 3.82–3.83 (m,
4H), 6.76 (s, 4H), 7.77 (d, J = 8.3 Hz, 4H), 8.01–8.06 (m, 8H).
6,16-Dimethyl-3,6,9,13,16,19-hexaaza-1,11(2,7)-dinaphthalenacyclo-
icosaphane (2,7-BisNP-NMe) was obtained from 5a in a 76% yield.
White powder, 1H NMR (300 MHz, CDCl3): δ 2.20 (s, 3H), 2.48 (t, 4H, J =
6.0 Hz), 2.73 (t, J = 5.7 Hz, 4H), 3.86 (s, 4H), 7.36 (dd, J = 8.4 Hz, 2H),
7.49 (s, 2H), 7.60 (d, J = 8.4 Hz, 1H). 2,7-BisNP-NMe × 6 HCl: The free-
base macrocycle (269 mg, 0.5 mmol) was dissolved in a mixture of
MeOH and 1,4-dioxane (2:1, 15 mL). HCl (4 M in 1,4-dioxane, 3.13 mL)
was added, and a precipitate has formed immediately. The volatiles were
removed in vacuo, and the residue was recrystallized from 0.5 M aq. HCl,
to give 2,7-BisNP-NMe × 6 HCl (476 mg, 0.58 mmol, quant. yield) as a
colorless crystalline solid. 1H NMR (300 MHz, 0.01 M DCl in D2O): δ 2.88
(s, 3H), 3.52 (s, 8H), 4.47 (s, 4H), 7.62 (d, J = 8.4 Hz, 2H), 8.06 (d, J =
8.5 Hz, 2H), 8.10 (s, 2H); 13C NMR (75 MHz, 0.01 M DCl in D2O): δ 41.4
(CH3), 41.8 (CH2), 51.9 (CH2), 52.5 (CH2), 128.5 (CH), 128.9 (Cq), 129.9
(CH), 130.9 (CH), 133.1 (Cq), 134.1 (Cq); MS (ESI+): m/z (%) 539.5 (100)
[M + H]+; anal. calcd. for C34H52N6Cl6 × 3.5 H2O (820.6): C, 49.76; H
7.25; N, 10.24; found C, 49.60; H, 7.03; N, 10.24.
1
General procedure for the synthesis of polyazamacrocyles: To a
suspension of the tetraimine intermediate (0.18 mmol) in a mixture of
CH2Cl2 (3.5 mL) and MeOH (1.5 mL), NaBH4 (104 mg, 2.76 mmol) was
added in one portion. The reaction mixture was stirred overnight at room
temperature and the solvents were removed under reduced pressure. To
the residue, aq. NaOH (1M, 30 mL) was added, and the mixture was
extracted with CH2Cl2 (3 x 20 mL). The combined organic layers were
washed with saturated K2CO3 solution (20 mL), dried with anhyd. K2CO3
and evaporated in vacuo. The residue was flash-chromatographed using
gradient elution with NH4OH (SiO2; eluent: CH2Cl2 / MeOH / aq. NH4OH,
80:20:2 to 80:20:4), to give the macrocyclic polyamine as a free base.
Novel macrocycles were fully characterized as hydrochloride salts.
3,9,13,19-Tetramethyl-6,16-dioxa-3,9,13,19-tetraaza-1,11(2,7)dinaph-
thalenacycloicosaphane (2,7-BisNP-O4Me): To a solution of 2,7-
BisNP-O (200 mg, 0.39 mmol) in HCOOH (0.82 mL), aq. HCHO (37%,
0.82 mL, 7.80 mmol) was added. The resulting mixture was heated at
120 °C for 24 h, evaporated under reduced pressure, and aq. NaOH (1M,
40 mL) was added. The aqueous layer was extracted with CH2Cl2 (3 x
25 mL), and the combined organic layers were dried with anhyd. K2CO3
and evaporated in vacuo. The residue was flash-chromatographed using
gradient elution with NH4OH (SiO2; eluent: CH2Cl2 / MeOH / aq. NH4OH,
80:20:0 to 80:20:2), to give the product as a white solid (140 mg,
0.25 mmol, 63%). 1H NMR (300 MHz, CDCl3): δ 2.28 (s, 3H), 2.59 (t, J =
5.6 Hz, 2H), 3.52 (t, J = 5.6 Hz, 2H), 3.63 (s, 2H), 7.39 (d, J = 8.4 Hz, 1H),
7.57 (s, 1H), 7.65 (d, J = 8.4 Hz, 1H). 2,7-BisNP-O4Me × 4 HCl: The
free-base macrocycle (281 mg, 0.5 mmol) was dissolved in a mixture of
MeOH and 1,4-dioxane (2:1, 10 mL). HCl (4 M in 1,4-dioxane, 2 mL) was
added and the solution was stirred until the hydrochloride salt
precipitated. The volatiles were evaporated under reduced pressure, and
the residue was purified by recrystallization from anhydrous isopropanol,
to give the product (214 mg, 0.30 mmol, 60%) as a white solid. 1H NMR
(300 MHz, D2O, 60 °C): δ 3.23 (s, 3H), 3.84 (m, 2H), 4.26 (m, 2H), 4.51
(m, 2H), 7.65 (d, J = 8.0 Hz, 1H), 8.18 (s, 1H), 8.23 (d, J = 8.5 Hz, 1H);
13C NMR (75 MHz, D2O, 60 °C): δ 41.7 (CH3), 55.4 (CH2), 61.1 (CH2),
65.3 (CH2), 128.0 (Cq), 129.0 (CH), 129.5 (CH), 131.5 (CH), 132.6 (Cq),
133.8 (Cq); MS (ESI+): m/z (%) 569.5 (100) [M + H]+; anal. calcd. for
C36H52N4O2Cl4 × 1.5 H2O (741.64): C, 58.30; H, 7.47; N, 7.55; found : C,
58.38; H, 7.41; N, 7.31.
6,16-Dioxa-3,9,13,19-tetraaza-1,11(2,7)-dinaphthalenacycloicosa-
phane (2,7-BisNP-O)[41] was obtained from 4 in a 88% yield. White
powder, 1H NMR (300 MHz, CDCl3): δ 2.84 (t, J = 5.0 Hz, 2H), 3.59 (t,
2H), 3.90 (s, 2H), 7.37 (d, J = 8.6 Hz, 1H), 7.63–7.65 (s + d, 2H).
6,16-Bis(3-hydroxypropyl)-3,6,9,13,16,19-hexaaza-1,11(2,7)-dinaph-
thalenacycloicosaphane (2,7-BisNP-NOH) was obtained from 5c in a
28% yield. Yellow oil, 1H NMR (300 MHz, CD3OD): δ 1.76 (br s, 2H),
2.70–2.76 (m, 6H), 2.92–2.96 (m, 4H), 3.62 (t, 2H), 3.90 (s, 4H), 7.30–
7.38 (m, 2H), 7.43–7.48 (m, 2H), 7.59–7.70 (m, 2H). 2,7-BisNP-NOH ×
6 HCl: The free-base macrocycle (31.9 mg, 0.05 mmol) was dissolved in
a mixture of MeOH and 1,4-dioxane (2:1, 1.2 mL). HCl (4 M in 1,4-
dioxane, 0.25 mL) was added, and the solution was stirred until the
hydrochloride salt precipitated. The volatiles were removed under
reduced pressure, to give 2,7-BisNP-NOH × 6 HCl (30.8 mg, 0.04 mmol,
71%) as a white powder. 1H NMR (300 MHz, D2O): δ 1.63 (quint, 2H),
2.58 (t, 2H), 2.81 (t, J = 5.3 Hz, 4H), 3.16 (t, J = 5.8 Hz, 4H), 3.52 (t, J =
5.8 Hz, 2H), 4.17 (s, 4H), 7.51 (d, J = 8.6 Hz, 2H), 7.89 (s, 2H), 7.99 (d, J
= 8.6 Hz, 2H); 13C NMR (75 MHz, D2O): δ 26.0 (CH2), 41.7 (CH2), 48.5
(CH2), 50.8 (CH2), 51.2 (CH2), 58.8 (CH2), 127.8 (CH), 128.2 (Cq), 129.1
(CH), 130.2 (CH), 132.3 (Cq), 133.3 (Cq); MS (ESI+): m/z (%) 627.5 (40)
[M + H]+; anal. calcd. for C38H60Cl6N6O2 × 3.8 H2O (914.04): C, 49.93; H,
7.45; N, 9.19; O, 10.15; found : C, 50.06, H, 7.16; N, 8.89; O, 9.88.
6,16-Bis(3-(tert-butoxycarbonyl)aminopropyl)-3,6,9,13,16,19-hexa-
3,3,9,9,13,13,19,19-Octamethyl-6,16-dioxa-3,9,13,19-tetraazonia-1,11
(2,7)-dinaphthalenacycloicosaphane iodide (2,7-BisNP-O-Me8, 4 I–):
In a sealed tube, the macrocycle 2,7-BisNP-O (205 mg, 0.4 mmol),
iodomethane (2.00 mL, 4.54 g, 32 mmol) and DIPEA (0.79 mL, 619 mg,
4.8 mmol) in DMF (12 mL) were stirred at 60 °C for 24 h. After cooling to
room temperature, the precipitate was collected, washed with DMF, dried
in vacuo, and recrystallized from water to give the product as a white
aza-1,11(2,7)-dinaphthalenacycloicosaphane
(2,7-BisNP-NNHBoc)
was obtained from 5b in a 67% yield. Colorless oil, 1H NMR (300 MHz,
CDCl3): δ 1.41 (s, 9H), 1.66 (br s, 2H), 2.40 (t, 2H), 2.52 (t, J = 5.1 Hz,
4H), 2.69 (t, J = 5.1 Hz, 4H), 3.07–3.08 (m, 2H), 3.86 (s, 4H), 5.85 (br s,
NH), 7.36 (d, J = 8.4 Hz, 2H), 7.47 (s, 2H), 7.65 (d, J = 8.4 Hz, 2H). 2,7-
BisNP-NNH2 × 8 HCl: The free-base macrocycle (149 mg, 0.18 mmol)
was dissolved in a mixture of MeOH and 1,4-dioxane (2:1, 4 mL). HCl (4
M in 1,4-dioxane, 0.8 mL) was added, and the solution was stirred until
the hydrochloride salt precipitated. The volatiles were removed under
reduced pressure, and the residue was purified by recrystallization from
EtOH–H2O to give 2,7-BisNP-NNH2 × 8 HCl as a white solid (87.7 mg,
1
solid (378 mg, 0.33 mmol, 83%). H NMR (300 MHz, [D6]DMSO:) δ 3.11
(s, 12H), 3.67 (s, 4H), 4.08 (s, 4H), 4.80 (s, 4H), 7.72 (d, J = 8.4 Hz, 2H),
8.09 (d, J = 8.4 Hz, 2H), 8.34 (s, 2H); 13C NMR (75 MHz, [D6]DMSO): δ
50.2 (CH3), 63.4 (CH2), 64.1 (CH2), 67.4 (CH2), 126.2 (Cq), 128.3 (CH),
131.1 (CH), 131.9 (Cq), 133.6 (CH), 134.1 (Cq); MS (ESI+): m/z (%) 967.6
(24) [M + 3 CF3COO–]+, 427.3 (75) [M + 2 CF3COO–]2+; anal. calcd. for
1
0.10 mmol, 53%). H NMR (300 MHz, D2O): δ 1.90–2.01 (m, 2H), 2.90–
3.03 (m, 4H), 3.13 (t, J = 6.4 Hz, 4H), 3.29 (m, 4H), 4.22 (s, 4H), 7.51 (d,
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
C40H60N4O2I4 × H2O (1154.58): C, 41.61; H, 5.41; N, 4.85; found: C,
41.77; H, 5.72; N, 4.48.
methods using SHELXS-97[62] and refined against F2 by full-matrix least-
squares techniques using SHELXL-2017[63] with anisotropic displacement
parameters for all non-hydrogen atoms. Hydrogen atoms were located on
a difference Fourier map and introduced into the calculations as a riding
model with isotropic thermal parameters. All calculations were performed
by using the Crystal Structure crystallographic software package
WINGX.[64] The crystal data collection and refinement parameters are
given in Supporting Information (Table S1). CCDC-1834477 & 1834478
contain the supplementary crystallographic data for this paper. These
data are provided free of charge by the Cambridge Crystallographic Data
Centre.
3,9,13,19-Tetra(N,N′-bis(tert-butoxycarbonyl)carbamimidoyl)-6,16-di-
oxa-3,9,13,19-tetraaza-1,11(2,7)dinaphthalenacycloicosaphane (6):
To a solution of 2,7-BisNP-O (512 mg, 1.0 mmol) in CH2Cl2 (50 mL), 1,3-
bis(tert-butoxycarbonyl)-2-methyl-2-thiopseudourea (1.45 g, 5.0 mmol),
NEt3 (2.78 mL, 20 mmol) and mercury(II) chloride (1.36 g, 5 mmol) were
added. The resulting mixture was stirred for 2 days at room temperature
and then filtered through a bed of Celite. The filtrate was washed with
water (3 × 50 mL), brine (50 mL), dried with MgSO4 and evaporated in
vacuo. The residue was flash-chromatographed using gradient elution
(SiO2; eluent: 0 to 10% MeOH in CH2Cl2), to give the product as a white
Nucleic acids: Oligonucleotides (lyophilized, RP-HPLC purity grade)
were purchased from Eurogentec, reconstituted in water at a
1
powder (1.24 g, 0.84 mmol, 84%). H NMR (300 MHz, CDCl3): δ 1.51 (s,
18H), 3.37 (br s, 4H), 4.78 (s, 2H), 7.37–7.46 (m, 2H), 7.74 (d, J = 8.3 Hz,
concentration of 100 µM, and stored at −20 °C. The equimolar amounts
of complementary strands were mixed in the adequate buffer and the
solutions were annealed for 5 min at 80 °C, slowly cooled to ambient
1H), 9.42 (br s, NH); MS (ESI+): m/z (%) 741.5 (100) [M + 2H]2+
.
temperature to give the duplex. Calf thymus DNA solution (10 mg mL−1
,
3,9,13,19-Tetra(carbamimidoyl)-6,16-dioxa-3,9,13,19-tetraaza-5,22-
1,11(2,7)dinaphthalenacycloicosaphane hydrochloride (2,7-BisNP-
O4Gua × 4 HCl): To a solution of compound 6 (1.24 g, 0.839 mmol) in
chloroform (4 mL), HCl (4 M in 1,4-dioxane, 4.20 mL, 16.8 mmol) was
added, the resulting mixture was stirred at room temperature during 24 h.
The volatiles were evaporated in vacuo and the product was purified by
two recrystallizations, first from EtOH and then from aq. 15% HCl, to give
the product as white crystals (270 mg, 0.31 mmol, 37%). 1H NMR (300
MHz, D2O): δ 3.53–3.54 (m, 4H), 3.61–3.62 (m, 4H), 4.66 (s, 4H), 7.35 (d,
J = 8.5 Hz, 2H), 7.44 (s, 2H), 7.86 (d, J = 8.5 Hz, 2H); 13C NMR (75 MHz,
D2O): δ 50.7 (CH2), 53.7 (CH2), 69.0 (CH2), 124.6 (CH), 125.5 (CH),
129.4 (CH), 132.6 (Cq), 133.4 (Cq), 134.1 (Cq), 157.9 (Cq); MS (ESI+): m/z
(%) 681.5 (100) [M + H]+; anal. calcd. for C36H52N12O2Cl4 × 2.5 H2O
(869.31): C, 49.60; H, 6.59; N, 19.28; found: C, 49.51; H, 6.31; N, 19.85.
Invitrogen) was diluted in a 10 mM KAsO2Me2, 10 mM KCl buffer (pH 7.2)
to a nucleotide concentration of ~5 mM (~50-fold), and the actual
concentration was determined by spectrophotometry, using the extinction
coefficient value of 12,824 cm−1 M−1 (base pairs, bp) at 260 nm. Solutions
were stored at +4 °C.
Fluorescence-melting experiments: The experiments were performed
with 17-mer oligonucleotides 17-NΦ (5′-(6-FAM)-CCAGTTCGNAGTAAC-
CC-3′ / 5′- GGGTTACTΦCGAACTGG-TAMRA-3′, where N = A, T, G or C,
Φ = THF / dSpacer), hybridized in a 10 mM KAsO2Me2, 10 mM KCl buffer
(pH 7.2) at a concentration of 25 µM. The samples (total volume: 25 µL)
were prepared by mixing duplex oligonucleotides (final concentration: 0.5
µM) and ligands (final concentration: 1 µM), in the absence or in the
presence of ct DNA competitor (final concentration: 60 or 200 µM bp) in a
10 mM KAsO2Me2, 10 mM KCl buffer (pH 7.2). Thermal denaturation
profiles were recorded in 96-well plates with a 7900HT Fast Real-Time
PCR apparatus (Applied Biosystems), using fluorescence detection in the
FAM channel. After an initial incubation at 25 °C for 5 min, the
temperature was increased to 95 °C in 0.5 °C increments every minute.
The temperatures of DNA melting transitions were determined from the
first-derivative plots of fluorescence intensity versus temperature. Each
experimental condition was tested in triplicate.
N,N-Bis(2-aminoethyl)-3-aminopropan-1-ol (7): To a mixture of N,N-
bis(2-phthalimidoethyl)amine (3.86 g, 10.6 mmol), K2CO3 (8.81 g, 63.8
mmol), KI (88 mg, 0.53 mmol) in MeCN (150 mL), 3-bromo-1-propanol
(5.54 mL, 8.86 g, 63.8 mmol) was added via syringe. The mixture was
heated under reflux overnight. The solvent was removed under reduced
pressure and H2O (100 mL) was added. The aqueous layer was
extracted with CH2Cl2 (3 × 50 mL). The combined organic layers were
washed with water, dried with MgSO4, filtered and concentrated under
reduced pressure. To the residue, aq. HCl (6 M, 6 mL) was added, and
the resulting mixture was heated under reflux overnight and then filtered.
The filtrate was stirred in presence of powdered charcoal during 30 min
and filtered again; the volatiles were removed under reduced pressure,
and MeOH (10 mL) was added. The solution was stirred in presence of
ion exchange resin (Amberlite IRA-402, OH– form, 15 mL), filtered, and
the solvent was removed under reduce pressure, to give the product 7 as
a yellow oil (400 mg, 2.48 mmol, 23%). 1H NMR (300 MHz, CD3OD): δ
1.73 (m, 2H), 2.53–2.62 (m, 6H), 2.73 (t, J = 6.3 Hz, 4H), 3.62 (t, J = 6.2
Hz, 2H); 13C NMR (75 MHz, CD3OD): δ 30.6 (CH2), 39.9 (CH2), 52.3
(CH2), 56.9 (CH2), 61.3 (CH2); MS (ESI+): m/z (%) 162.2 (100) [M + H]+.
Fluorimetric APE1 inhibition assay: Recombinant, polyHis-tagged
APE1 was purchased from Life Technologies, reconstituted in a buffer
containing 10 mM Tris-HCl, 50 mM NaCl, 1 mM DTT, 0,05 mM Na2EDTA,
200 µg mL-1 BSA, and 50 % glycerol (pH 8.0) at a concentration of
100 nM, aliquoted and stored at −20 °C. The substrate 17-TΦ (5′-
ATTO390-CCAGTTCGTAGTAACCC-3′ / 5′-GGGTTACTΦCGAACTGG-
DABCYL-3′) was hybridized in a 250 mM KAsO2Me2, 250 mM KCl (pH
7.2) buffer at a concentration of 20 µM. For kinetic studies, samples of
17-TΦ (final concentration: 0.2 µM, total volume: 1.00 mL) in the absence
or in the presence of varied concentrations of ligands were diluted in the
APE1 reaction buffer (50 mM HEPES, 50 mM NaCl, 1 mM MgCl2, 2 mM
DTT, pH 8.0), thermostated at 37 °C in disposable acrylic semi-micro
cuvettes (Evergreen Scientific) for at least 5 min, and an aliquot of APE1
was added. The amount of APE1 was adjusted in order to maintain the
same APE1 activity in the absence of inhibitors (v0 ≈ 0.02 s−1). The real-
time fluorescence readings were performed every 12 s with an Agilent
Cary Eclipse spectrofluorimeter running the following experimental para-
meters: λex = 395 nm; λem = 465 nm; slit widths = 10 nm; PMT voltage =
700 V, T = 37 °C; run duration = 90 s. In order to account for the partial
quenching effect induced by some ligands, all fluorescence intensity plots
were normalized with respect to the initial intensity (F0, before addition of
the enzyme), and initial reaction velocities (v0) were determined by
manual fitting of linear regions of the curves F/F0 = ƒ(t) (r2 > 0.9). Each
Single-crystal X-ray crystallography: X-ray diffraction data for 2,7-
BisNP-O4Gua × 4 HCl were collected with a Bruker VENTURE /
PHOTON 100 CMOS diffractometer with Micro-focus IµS source Mo Kα
radiation. X-ray diffraction data for 2,7-BisNP-Im (2 Br−) were collected
with
a Bruker X8 APEX II CCD diffractometer with graphite-
monochromated Mo-Kα radiation. Crystals were mounted on a CryoLoop
(Hampton Research) with Paratone-N (Hampton Research) as
cryoprotectant and then flash-frozen in a nitrogen-gas stream at 100 K.
For both compounds, the temperature of the crystal was maintained at
the selected value by means of a 700 series Cryostream cooling device
to within an accuracy of ±1 K. The data were corrected for Lorentz
polarization, and absorption effects. The structures were solved by direct
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
experimental condition was tested in triplicate. The enzymatic activity
10 mM KAsO2Me2 buffer, pH 7.2 (total volume: 1000 µL) overnight at
37 °C. Then, the mixtures were concentrated to ~150 µL, cold EtOH
(900 µL) was added and samples were stored overnight at −20 °C. After
centrifugation, the supernatant was removed and the pellet was dried,
than resuspended in 0.1 M TEAA (10 µL). Excess of salt was removed
using ZipTip-C18 (Millipore) following the manufacturer’s protocol. LC/MS
analysis was performed on a Waters BEH C18 column (130 Å, 1.7 μm,
2.1 × 50 mm), installed on a Waters ACQUITY H-Class UPLC system
with a SQ Detector 2 mass detector. Eluent A: 15 mM tri-n-butylamine +
50 mM HFIP, eluent B: MeOH, flow: 0.2 mL min−1, gradient: 0 to 1 min,
0–15% B (linear); 1 to 3.5 min, 15–60% B (linear); 3.5 to 4.5 min, 0.1%
formic acid; 4.5 to 6.0 min, 100% A (re-equilibration step). Data were
analyzed using Waters MassLynx software.
was calculated as a ratio of v0 in the presence (v0inh) and in the absence
inh
of inhibitor (% activity = v0 / v0) and plotted as a function of the
logarithm of concentration of inhibitor. IC50 values were determined by
fitting these plots with a dose-response function.
Real-time fluorimetric assay for AP-DNA cleavage: The double-
labelled uracil-containing precursor 17-TU (5′-ATTO390-CCAGTT-
CGTAGTAACCC-3′
/ 5′-GGGTTACTUCGAACTGG-DABCYL-3′) was
hybridized in a 250 mM KAsO2Me2, 250 mM KCl buffer (pH 7.2) at a
concentration of 40 µM. The batches of the abasic substrate 17-TX were
prepared, on a daily basis, by incubation of 17-TU (final concentration:
20 µM) with UDG (NEB, 5000 U mL−1, 1 µL) in 1X UDG Reaction Buffer
(20 mM Tris-HCl, 1 mM DTT, 1 mM EDTA, pH 8, total reaction volume:
120 µL) for 45 min at 37 °C. To determine the rates of ligand-induced
substrate cleavage, samples of 17-TX (final concentration: 0.2 µM in
1.00 mL of a 10 mM KAsO2Me2, 150 mM KCl buffer, pH 7.2) were
thermostatted at 37 °C in disposable acrylic semi-micro cuvettes
(Evergreen Scientific) and, after 5 min, aliquots of ligands were added to
a final concentration of 5 µM. The real-time fluorescence readings were
performed every 12 s with an Agilent Cary Eclipse spectrofluorimeter
Acknowledgements
The authors thank Dr. Marie-Paule Teulade-Fichou for helpful
and stimulating discussions, Dr. Marcel Hollenstein and Dr. Ivo
Sarac (Institut Pasteur, Paris) for help with LC/MS analysis of
oligonucleotides, Ms. Séverine Soille, Ms. Flore Migeon and Ms.
Isciane Commenge for help with biochemical experiments, and
Dr. Michela Zuffo for careful reading of the manuscript. This
work was supported by the “IDI 2016” project funded by the
IDEX Paris-Saclay (ANR-11-IDEX-0003-02) through a PhD
scholarship to C.C.
running the following experimental parameters: λex = 395 nm; λem
=
465 nm; slit widths = 5 nm; PMT voltage = 700 V; T = 37 °C; run duration
= 180 min. The initial slopes (v0) of the fluorescence intensity vs. time
plots were determined by manual fitting of the linear regions of the curves
(r2 > 0.9). Each experimental condition was tested in triplicate.
To account for the extent of ligand-induced fluorescence quenching, the
above-described experiments were performed with a non-cleavable THF
analogue 17-TΦ (5′-ATTO390-CCAGTTCGTAGTAACCC-3′ / 5′-GGGTT-
ACTΦCGAACTGG-DABCYL-3′), and quenching factors were calculated
as a ratio of time-averaged fluorescence before and after addition of
ligands (QF = Fwith ligand / Fno ligand).
Keywords: Abasic sites • APE1 • DNA cleavage • macrocycles •
DNA repair
[1]
[2]
J. Lhomme, J. F. Constant, M. Demeunynck, Biopolymers 1999, 52,
65–83.
Gel electrophoresis experiments: The uracil-containing strand (5′-
GGGTTACTUCGAACTGG-3′) was 5′-32P-labelled according to standard
P. Fortini, B. Pascucci, E. Parlanti, M. D’Errico, V. Simonelli, E. Dogliotti,
Biochimie 2003, 85, 1053–1071.
protocols and purified on
a Sephadex column. The radiolabelled
precursor 17-TU was prepared by annealing two complementary non-
labelled strands (40 µM) in the presence of a small amount of the 5′-32P-
labelled strand (~2 × 106 cpm), and the abasic substrate 17TX was
subsequently obtained by UDG treatment as described above. The
radiolabelled substrate (final concentration: 0.2 µM) was incubated at
37 °C for 1 hour (unless stated otherwise) in the absence or in the
presence of ligands (final concentration: 5 µM) in 10 mM KAsO2Me2, 150
mM KCl buffer, pH 7.2 (total volume: 20 µL). Gel loading buffer
(formamide / 1 M EDTA / bromophenol blue / xylene cyanol, 12 µL) was
added, and samples (aliquots of 8 µL) were loaded onto a denaturing (7
M urea), 20% polyacrylamide gel and separated at ambient temperature
during 2.5 h (~1500 V, 20 W). After exposure to a phosphorimager
screen overnight, the bands were visualized with a Typhoon Trio imager
(GE Healthcare) and quantified with ImageQuant TL software (GE
Healthcare). Each experimental condition was tested in triplicate. The
yields of strand cleavage products were calculated relative to total
radioactivity in each lane.
[3]
[4]
[5]
K. S. Gates, Chem. Res. Toxicol. 2009, 22, 1747–1760.
D. Fu, J. A. Calvo, L. D. Samson, Nat. Rev. Cancer 2012, 12, 104–120.
K. Yang, D. Park, N. Y. Tretyakova, M. M. Greenberg, Proc. Natl. Acad.
Sci. 2018, in press (doi: 10.1073/pnas.1813338115).
J. Zhang, M. F.G. Stevens, T. D. Bradshaw, Curr. Mol. Pharmacol.
2012, 5, 102–114.
[6]
[7]
[8]
[9]
M. S. Bobola, D. D. Kolstoe, A. Blank, M. C. Chamberlain, J. R. Silber,
Front. Oncol. 2012, 2, 1–9.
H. Erasimus, M. Gobin, S. Niclou, E. Van Dyck, Mutat. Res. - Rev.
Mutat. Res. 2016, 769, 19–35.
G. Damia, M. D’Incalci, Eur. J. Cancer 2007, 43, 1791–1801.
[10] Y. Zhu, J. Hu, Y. Hu, W. Liu, Cancer Treat. Rev. 2009, 35, 590–596.
[11] R. I. Al-Safi, S. Odde, Y. Shabaik, N. Neamati, Curr. Mol. Pharmacol.
2012, 5, 14–35.
[12] G. Kaur, R. P. Cholia, A. K. Mantha, R. Kumar, J. Med. Chem. 2014, 57,
10241–10256.
[13] S. S. Laev, N. F. Salakhutdinov, O. I. Lavrik, Bioorg. Med. Chem. 2017,
25, 2531–2544.
LC/MS analysis of covalent adduct: The uracil-containing precursor
17-TU (5′-CCAGTTCGTAGTAACCC-3′ / 5′-GGGTTACTUCGAACTGG-
3′) was annealed in a 250 mM KAsO2Me2, 250 mM KCl buffer (pH 7.2) at
a concentration of 40 µM. The batches of the abasic substrate 17-TX
were prepared by incubation of 17-TU (final concentration: 20 µM) with
UDG (NEB, 5000 U mL−1, 1 µL) in 1X UDG Reaction Buffer (20 mM Tris-
HCl, 1 mM DTT, 1 mM EDTA, pH 8, total reaction volume: 120 µL) for 45
min at 37 °C. Samples of 17-TX (final concentration: 0.2 µM) and 2,7-
BisNP-NH (final concentration: 50 µM) were incubated in 150 mM KCl,
[14] A. P. Montaldi, E. T. Sakamoto-Hojo, Clin. Exp. Med. 2013, 13, 279–88.
[15] S. Wei, S. Shalhout, Y. H. Ahn, A. S. Bhagwat, DNA Repair 2015, 27,
9–18.
[16] NCT02395692, Methoxyamine and Temozolomide in treating patients
with recurrent glioblastoma. Bethesda (MD), National Library of
Medicine (US); https://clinicaltrials.gov/ct2/show/NCT02395692, 2018.
[17] P. Belmont, A. Boudali, J. F. Constant, M. Demeunynck, A. Fkyerat, P.
Michon, G. Serratrice, J. Lhomme, New J. Chem. 1997, 21, 47–54.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
[18] B. M. Zeglis, J. A. Boland, J. K. Barton, Biochemistry 2009, 48, 839–
849.
[42] M. Jourdan, A. Granzhan, R. Guillot, P. Dumy, M.-P. Teulade-Fichou,
Nucleic Acids Res. 2012, 40, 5115–5128.
[19] K. Benner, A. Granzhan, H. Ihmels, G. Viola, Eur. J. Org. Chem. 2007,
2007, 4721–4730.
[43] N. Kotera, R. Guillot, M.-P. Teulade-Fichou, A. Granzhan,
ChemBioChem 2017, 18, 618–622.
[20] K. Benner, A. Bergen, H. Ihmels, P. M. Pithan, Chem. Eur. J. 2014, 20,
9883–9887.
[44] C. Yung-Chi, W. H. Prusoff, Biochem. Pharmacol. 1973, 22, 3099–3108.
[45] G. Rai, V. N. Vyjayanti, D. Dorjsuren, A. Simeonov, A. Jadhav, D. M.
Wilson, D. J. Maloney, J. Med. Chem. 2012, 55, 3101–12.
[46] M. C. Rideout, B. Liet, D. Gasparutto, N. Berthet, Anal. Biochem. 2016,
513, 93–97.
[21] K. Benner, H. Ihmels, S. Kölsch, P. M. Pithan, Org. Biomol. Chem.
2014, 12, 1725–34.
[22] J. Malina, P. Scott, V. Brabec, Nucleic Acids Res. 2015, 43, 5297–5306.
[23] N. Berthet, J. Michon, J. Lhomme, M.-P. Teulade-Fichou, J.-P.
Vigneron, J.-M. Lehn, Chem. Eur. J. 1999, 5, 3625–3630.
[24] M. Jourdan, J. Garcia, J. Lhomme, M.-P. Teulade-Fichou, J.-P.
Vigneron, J.-M. Lehn, Biochemistry 1999, 38, 14205–14213.
[25] N. Kotera, F. Poyer, A. Granzhan, M.-P. Teulade-Fichou, Chem.
Commun. 2015, 51, 15948–15951.
[47] R. Male, V. M. Fosse, K. Kleppe, Nucleic Acids Res. 1982, 10, 6305–
6318.
[48] A. Mazumder, J. A. Gerlt, M. J. Absalon, J. A. Stubbe, R. P.
Cunningham, J. Withka, P. H. Bolton, Biochemistry 1991, 30, 1119–
1126.
[49] R. Rahimoff, O. Kosmatchev, A. Kirchner, T. Pfaffeneder, F. Spada, V.
Brantl, M. Müller, T. Carell, J. Am. Chem. Soc. 2017, 139, 10359–
10364.
[26] N. Kotera, A. Granzhan, M.-P. Teulade-Fichou, Biochimie 2016, 128–
129, 133–137.
[27] J. Pierre, J. Laval, J. Biol. Chem. 1981, 256, 10217–10220.
[28] T. Behmoaras, J. J. Toulmé, C. Hélène, Nature 1981, 292, 858–859.
[29] A. J. Kurtz, M. L. Dodson, R. S. Lloyd, Biochemistry 2002, 41, 7054–
7064.
[50] K. M. Johnson, N. E. Price, J. Wang, M. I. Fekry, S. Dutta, D. R. Seiner,
Y. Wang, K. S. Gates, J. Am. Chem. Soc. 2013, 135, 1015–1025.
[51] N. E. Price, K. M. Johnson, J. Wang, M. I. Fekry, Y. Wang, K. S. Gates,
J. Am. Chem. Soc. 2014, 136, 3483–3490.
[30] A. Fkyerat, M. Demeunynck, J. F. Constant, P. Michon, J. Lhomme, J.
Am. Chem. Soc. 1993, 115, 9952–9959.
[52] Y. Zhiyu, N. E. Price, K. M. Johnson, W. Yinsheng, K. S. Gates, Nucleic
Acids Res. 2017, 45, 6275–6283.
[31] V. Steullet, S. Edwards-Bennett, D. W. Dixon, Bioorganic Med. Chem.
1999, 7, 2531–2540.
[53] T. Marsoner, O. P. Schmidt, T. Triemer, N. W. Luedtke, ChemBioChem
2017, 18, 894–898.
[32] M. Perigolo De Oliveira, J. F. Constant, M. Peuchmaur, I. Pitta, J. L.
Décout, Chem. Res. Toxicol. 2013, 26, 1710–1719.
[54] A. C. Komor, C. J. Schneider, A. G. Weidmann, J. K. Barton, J. Am.
Chem. Soc. 2012, 134, 19223–33.
[33] I. G. Minko, A. C. Jacobs, A. R. De Leon, F. Gruppi, N. Donley, T. M.
Harris, C. J. Rizzo, A. K. McCullough, R. S. Lloyd, Sci. Rep. 2016, 6,
28894.
[55] K. M. Boyle, J. K. Barton, J. Am. Chem. Soc. 2018, 140, 5612–5624.
[56] D. Melton, C. D. Lewis, N. E. Price, K. S. Gates, Chem. Res. Toxicol.
2014, 27, 2113–2118.
[34] Y. S. Abe, S. Sasaki, Bioorganic Med. Chem. 2016, 24, 910–914.
[35] J. M. Barret, C. Etievant, J. Fahy, J. Lhomme, B. T. Hill, Anticancer
Drugs 1999, 10, 55–65.
[57] J. T. Sczepanski, R. S. Wong, J. N. McKnight, G. D. Bowman, M. M.
Greenberg, Proc. Natl. Acad. Sci. 2010, 107, 22475–22480.
[58] J. T. Sczepanski, C. Zhou, M. M. Greenberg, Biochemistry 2013, 52,
2157–2164.
[36] S. N. Andres, M. J. Schellenberg, B. D. Wallace, P. Tumbale, R. S.
Williams, Environ. Mol. Mutagen. 2015, 56, 1–21.
[59] M. M. Greenberg, Acc. Chem. Res. 2014, 47, 646–655.
[60] J. Novotna, A. Laguerre, A. Granzhan, M. Pirrotta, M.-P. Teulade-
Fichou, D. Monchaud, Org. Biomol. Chem. 2015, 13, 215–222.
[61] A. Granzhan, M.-P. Teulade-Fichou, Tetrahedron 2009, 65, 1349–1360.
[62] G. M. Sheldrick, SHELXS-97, Program for X-Ray Crystal Structure
Solution, University Of Gottingen, Germany, 1997.
[37] M. V. Baker, M. J. Bosnich, C. C. Williams, B. W. Skelton, A. H. White,
Aust. J. Chem. 1999, 52, 823.
[38] E. Alcalde, N. Mesquida, L. Pérez-García, C. Alvarez-Rúa, S. García-
Granda, E. García-Rodriguez, Chem. Commun. 1999, 295–296.
[39] A. Caballero, N. G. White, P. D. Beer, Angew. Chem. Int. Ed. 2011, 50,
1845–1848.
[63] G. M. Sheldrick, Acta Crystallogr. Sect. C Struct. Chem. 2015, 71, 3–8.
[64] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[40] F. Zapata, A. Caballero, N. G. White, T. D. W. Claridge, P. J. Costa, V.
Félix, P. D. Beer, J. Am. Chem. Soc. 2012, 134, 11533–11541.
[41] A. Granzhan, E. Largy, N. Saettel, M.-P. P. Teulade-Fichou, Chem. Eur.
J. 2010, 16, 878–889.
This article is protected by copyright. All rights reserved.

10.1002/chem.201805555
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Cationic naphthalenophanes bind
to abasic (AP) sites in double-
stranded DNA with high affinity and
selectivity. Depending on their
molecular structure, this process
may lead to efficient indirect
inhibition of enzymatic (APE1)
hydrolysis of AP-sites, ligand-
induced DNA cleavage, or formation
of covalent ligand–DNA adducts.
Coralie Caron, Xuan N. T. Duong,
Régis Guillot, Sophie Bombard, Anton
Granzhan*
APE1
APE1
Page No. – Page No.
APE1 inhibition
abasic
(AP) site
Interaction of Functionalized
Naphthalenophanes with Abasic
Sites in DNA: DNA Cleavage, DNA
Cleavage Inhibition, and Formation
of Ligand–DNA Adducts
+
AP-site cleavage
covalentadduct
This article is protected by copyright. All rights reserved.